摘要
根据决策表信息系统的分明矩阵及序贯思想,提出了序贯属性约简算法,该算法首先构造递增序列的分明函数,然后利用逐次增加的属性核对分明函数进行分支运算,并建立属性约简树,从而求出所有约简.该算法避免了大量的逻辑运算,实现了高维数据的高效属性约简.理论分析和实验结果表明该算法具有更高的运行效率.
This paper proposes sequential attribute reduction algorithm, which is based on combination of decision table information system＇s discernibility matrix and sequential idea, this method generates increasing discernibility function series, and branch operation is implemented using successive increase attribute core, to build up the attribute reduction tree, thereby all reductions are obtained. Plenty of logical calculus is protected in this algorithm, realize highly active attribute reduction for high dimension data. Theory analysis and the experimental results show this algorithm costs less time than other algorithms.
出处
《微电子学与计算机》
CSCD
北大核心
2009年第7期95-97,101,共4页
Microelectronics & Computer
基金
河南省重点攻关项目(082102210015)
河南科技大学青年基金项目(2007QN041)
关键词
粗糙集
序贯属性约简
分明矩阵
属性核
rough sets
sequential attribute reduction
discernibility matrix
attribute core
